1. Field of the Invention
The present invention relates to the technical field of image data processing and, more particularly, to a method for reducing the data amount in an image memory system.
2. Description of Related Art
Image compression and decompression are widely used in video and multimedia applications. In accordance with certain similarities between two adjacent frames of consecutive pictures, image compression can highly increase the efficiency of data compression. This is achieved by using motion estimation to omit data where a current frame is similar to the nearest preceding frame, and only recording the different ones. Such a compression method is also referred to as a frame-related compression. Image compression is performed in a unit of GOP (group of picture) consisting of three types, i.e. I, B and P frames. In general, a GOP has 15 frames: an I-frame at the first position, a P-frame at the second position, two B-frames at the next two positions, and so on. As such, the GOP is composed of a sequence of IPBB PBB PBB. I frame referred to as a full frame is an intra-coded picture which is coded independently without reference to other pictures and provides an access point to the coded sequence where decoding can begin, but is coded with only moderate compression; P frame is a predictive-coded picture which is obtained by omitting data of a current frame similar to the nearest preceding frame; and B frame is a bidirectionally predictive-coded picture which is obtained by omitting data of a current frame similar to that of the nearest preceding and upcoming pictures. The P and B frames are not complete frames that require reference to the I frame.
However, with current DRAM standards, DRAMs as image buffers can only have some specified capacities. When regarding cost, the data amount of each compressed frame (plus other required information) is still greater than the capacity of a typical DRAM. To reduce the data amount for meeting current DRAM standards, the compressed frames or pictures are further compressed. However, a conventional approach for further compression only processes a half frame of a motion picture at a time, thereby adding processing time and reducing the entire system efficiency. Therefore, it is desirable to provide an improved method to mitigate and/or obviate the aforementioned problems.